Dataset Information

Liner programming method to construct equated item sets for implementation of periodical computer-based testing of Korean medical licensing examination.

ABSTRACT: PURPOSE:This study aimed to find the best way of developing equivalent item sets and to propose a stable and effective management plan for the periodical licensing examinations. METHODS:Five pre-equated item sets were developed based on the predicted correct answer rate of each item by using linear programming. These pre-equated item sets were compared to the ones that were developed with random item selection method based on the actual answer rate and difficulty from item response theory (IRT). Also, the results with and without common items were compared in the same way. ACAR and the IRT difficulty was used to determine whether there is a significant difference between pre-equating conditions. RESULTS:There was a statistically significant difference in IRT difficulty among the results from different pre-equated conditions. As predicted correct answer rate was divided into 2 or 3 difficulty boundaries, the actual answer rate and IRT difficulty parameters of the 5 item sets were equally constructed. Comparing item sets conditions with common items and without common items, including common items did not contribute much for the equating of 5 item sets. CONCLUSION:This study suggested the linear programming method is applicable to construct equated-item sets that reflect each content area. The best method to construct equated item sets suggested is to divide the predicted correct answer rate into 2 or 3 difficulty boundaries regardless of common items. If pre-equated item sets are required to construct a test based on the actual data, several optimal methods should be considered by simulation studies before administrating a real test.

SUBMITTER: Seo DG

PROVIDER: S-EPMC6318553 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Liner programming method to construct equated item sets for implementation of periodical computer-based testing of Korean medical licensing examination.

Seo Dong Gi DG Kim Myeong Gi MG Kim Na Hui NH Shin Hye Sook HS Kim Hyun Jung HJ

Journal of educational evaluation for health professions 20181018

<h4>Purpose</h4>This study aimed to find the best way of developing equivalent item sets and to propose a stable and effective management plan for the periodical licensing examinations.<h4>Methods</h4>Five pre-equated item sets were developed based on the predicted correct answer rate of each item by using linear programming. These pre-equated item sets were compared to the ones that were developed with random item selection method based on the actual answer rate and difficulty from item respons ...[more]

PMID: 30336541

Similar Datasets

Project description:PurposeThe purpose of this study was to improve the quality of items on the Korean Nursing Licensing Examination by developing and evaluating case-based items that reflect integrated nursing knowledge.MethodsWe conducted a cross-sectional observational study to develop new case-based items. The methods for developing test items included expert workshops, brainstorming, and verification of content validity. After a mock examination of undergraduate nursing students using the newly developed case-based items, we evaluated the appropriateness of the items through classical test theory and item response theory.ResultsA total of 50 case-based items were developed for the mock examination, and content validity was evaluated. The question items integrated 34 discrete elements of integrated nursing knowledge. The mock examination was taken by 741 baccalaureate students in their fourth year of study at 13 universities. Their average score on the mock examination was 57.4, and the examination showed a reliability of 0.40. According to classical test theory, the average level of item difficulty of the items was 57.4% (80%-100% for 12 items; 60%-80% for 13 items; and less than 60% for 25 items). The mean discrimination index was 0.19, and was above 0.30 for 11 items and 0.20 to 0.29 for 15 items. According to item response theory, the item discrimination parameter (in the logistic model) was none for 10 items (0.00), very low for 20 items (0.01 to 0.34), low for 12 items (0.35 to 0.64), moderate for 6 items (0.65 to 1.34), high for 1 item (1.35 to 1.69), and very high for 1 item (above 1.70). The item difficulty was very easy for 24 items (below -2.0), easy for 8 items (-2.0 to -0.5), medium for 6 items (-0.5 to 0.5), hard for 3 items (0.5 to 2.0), and very hard for 9 items (2.0 or above). The goodness-of-fit test in terms of the 2-parameter item response model between the range of 2.0 to 0.5 revealed that 12 items had an ideal correct answer rate.ConclusionWe surmised that the low reliability of the mock examination was influenced by the timing of the test for the examinees and the inappropriate difficulty of the items. Our study suggested a methodology for the development of future case-based items for the Korean Nursing Licensing Examination.

Project description:Traditional Korean medicine (TKM) emphasizes individualized diagnosis and treatment. This uniqueness makes AI modeling difficult due to limited data and implicit processes. Large language models (LLMs) have demonstrated impressive medical inference, even without advanced training in medical texts. This study assessed the capabilities of GPT-4 in TKM, using the Korean National Licensing Examination for Korean Medicine Doctors (K-NLEKMD) as a benchmark. The K-NLEKMD, administered by a national organization, encompasses 12 major subjects in TKM. GPT-4 answered 340 questions from the 2022 K-NLEKMD. We optimized prompts with Chinese-term annotation, English translation for questions and instruction, exam-optimized instruction, and self-consistency. GPT-4 with optimized prompts achieved 66.18% accuracy, surpassing both the examination's average pass mark of 60% and the 40% minimum for each subject. The gradual introduction of language-related prompts and prompting techniques enhanced the accuracy from 51.82% to its maximum accuracy. GPT-4 showed low accuracy in subjects including public health & medicine-related law, internal medicine (2), and acupuncture medicine which are highly localized in Korea and TKM. The model's accuracy was lower for questions requiring TKM-specialized knowledge than those that did not. It exhibited higher accuracy in diagnosis-based and recall-based questions than in intervention-based questions. A significant positive correlation was observed between the consistency and accuracy of GPT-4's responses. This study unveils both the potential and challenges of applying LLMs to TKM. These findings underline the potential of LLMs like GPT-4 in culturally adapted medicine, especially TKM, for tasks such as clinical assistance, medical education, and research. But they also point towards the necessity for the development of methods to mitigate cultural bias inherent in large language models and validate their efficacy in real-world clinical settings.

Dataset Information

Liner programming method to construct equated item sets for implementation of periodical computer-based testing of Korean medical licensing examination.

Publications

Liner programming method to construct equated item sets for implementation of periodical computer-based testing of Korean medical licensing examination.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets